C. D. Warren writes: The authors are to be congratulated on several interesting observations but, with respect to the likely cause of settlement of the Dubai dry dock pier D caissons, it is important to note that their fig. 12 gives a totally simplistic and possibly distorted view of the actual geological conditions that exist beneath the settling caissons, as indicated by boreholes. It must be stated that at the levels which suffered settlement, i.e. -17mOD and to a lesser extent -21.5 mOD, the boreholes did encounter thin horizons of generally uncemented gypsiferous sandy silts, up to 1 m thick, within the more competent calcarenite rock, made up of alternating units of miliolite sandstone and bioclastic limestone (Warren 1985). Therefore a mechanical process should not be totally discounted on the information presently available, particularly in view of the fact that channelled water flow and ‘piping’ of similar material was noted during mapping the numerous excavations extending beneath the dock floor, see Fig. 1.
The uncemented fine sandy silts, present at the boundary between the bioclastic limestone and overlying miliolite sandstone, displayed a marked variation in thickness, degree of cementation and gypsum content over short lateral distances, believed to reflect their former environment of deposition and subsequent mode of formation. Such facies variations, together with occurance of a saline groundwater intrusion beneath the dock itself, could partly explain the difference in both rock and water chemistries quoted by the author as supporting evidence for gypsum solution. However, it is to be